The Most Underrated Use for AI During Meetings: Being Your Devil’s Advocate
Source: Freepik
When it comes to your meetings, you may think AI is useless. Your team already has structured brainstorming sessions where everyone contributes and treats others with respect.
But that harmony could be the exact reason that you need to experiment with AI. Your team might be getting along TOO well.
Discourse without disagreement can make for pleasant meetings and good team morale. However, truly innovative and disruptive ideas should be, well, disruptive.
Having at least one person (the devil's advocate) disagree with the rest of the team can often lead to one of two things:
The team successfully defends the consensus, and everyone can move forward confidently knowing it’s the right call.
The devil’s advocate successfully argues that the consensus is flawed. A new and better idea is reached, and the team moves forward knowing they possibly dodged a bullet.
The right AI can play that role for you. And here’s how.
The problem(s) with getting along too well
You’ve deliberately built a company culture of respect that attracts like-minded people. So why would you want to purposely add conflict? That’s not exactly what we’re describing. But...
If you don’t run into any sort of disagreement or dissent in your meetings, you run the risk of two major problems:
Groupthink: A psychological phenomenon that occurs when a group of people consciously or unconsciously prioritize a harmonious group dynamic over the intended outcome.
For example, someone doesn’t want to “rock the boat,” so they silently hold onto a divergent opinion or idea to keep things moving. As a result, Groupthink can hurt creativity and innovation.
Hindsight Bias: When something in the project goes wrong, the team looks back and says things like “We should have known that __ would have been a problem” or “Of course, ___ took longer than we estimated.” Hindsight is always 20/20.
This type of talk often pops up in agile retrospectives and can lead to a lot of unproductive finger-pointing and blame.
A little bit of disagreement during planning can save a major argument later during project postmortems or retrospectives.
Does your agile team struggle with unproductive retrospectives?
Find out how AI can help with our free eBook, Uncovering Blind Spots in Agile Retrospectives: Advancing Retrospectives with AI.
Let AI be your devil’s advocate
We’re not saying you should completely default to, or defer to, AI-generated ideas alone. And we’re definitely not saying you should give AI a seat at the decision-making table. We’re actually encouraging you to use it to take your brainstorming sessions to the next level, or unclog blocked minds when you reach a dead end.
Try this during your next brainstorming session or PI Planning session. Proceed normally and go as far as the collective minds in the meeting will take you. Once you’ve reached that point, bring AI into the discussion.
For example, if you’re a Stormboard user, press the AI button on a section after your team has filled that area with sticky notes full of ideas. Our contextual AI will generate new ideas and concepts related to that section and the data it sees elsewhere in your Storm.
If you’re not a Stormboard user, you can try using a Large Language Model (LLM) and asking it relevant questions.
In either case, introducing AI’s perspective will help you:
Strengthen Arguments: Getting your team to defend its position or idea is a good thing. It forces you to dig deeper and come up with concrete arguments against the alternative. As a result, you can be 100% sure your idea is the best one possible.
Reveal blind spots or weaknesses: AI is the perfect stress test for ideas. If the current idea has any weak spots, AI will help you uncover them with perspectives you didn’t think of.
Uncover biases: Sometimes you can’t see the forest for the trees, particularly when your team has been staring at the same whiteboard for hours. Or even weeks!
AI is a fresh set of eyes that looks beyond any subconscious biases that may have crept into your decision-making along the way. More on this in the next section.
Escape the Echo Chamber: We’ve all reached the point in the meeting where the group is just building on existing ideas. Everyone is tired and all they can muster is different versions of “same, but different.”
Why this works
A number of industry-leading teams have experimented with assigning one person to be the devil’s advocate for brainstorming sessions. This person is assigned to make the case for the opposite of a group consensus.
It was an innovative idea, but most found that it fell flat for two key reasons:
The group knew this person was just “playing” devil’s advocate: The team was keenly aware that the person was assigned this task. They had a tough time taking the devil’s advocate, or any counterpoints they brought up, seriously because they assumed this person was just playing their role.
The assigned devil’s advocate rarely made a compelling case: The assigned advocate was almost always from the existing team. This means that they were often a part of the team’s existing biases and rarely believed in the case they were preparing.
No matter how seriously they took the advocate assignment, it’s hard to passionately and convincingly make a case for something you don’t really believe.
As a result, most organizations found themselves without the fresh ideas or innovation they had hoped for.
Using AI can eliminate these problems. The team knows AI is not making a case against the consensus view for the sake of a creative exercise.
AI can eliminate the 9 unique biases that ruin decision-making
AI’s new perspectives can help to shine a light on any biases you didn’t know your team has.
This includes the 9 different types of biases that are known to cloud decision-making for both individuals and groups.
Authority bias
Defaulting to an authority figure’s ideas (high-ranking employee or executive) over others (lower-ranking team member) because we assume the source is more valuable.
Anchoring bias
Human beings have a natural tendency to gravitate toward the first piece of information we receive. Anything after that has to work much harder to be considered an option.
Availability bias
Much like anchoring bias, you’re convinced that the oldest information you have on the subject is the most accurate.
Self-serving bias
Unintentionally making decisions that will benefit you (or the organization) over external parties like vendors, partners, or even customers/clients.
Confirmation bias
You start with the desired outcome, then seek data to confirm existing beliefs while rejecting any other information that defies your desired conclusion/action.
Conformity bias
A symptom of Groupthink. An individual makes a decision based on the majority consensus.
Framing bias
The way information is presented impacts how valuable we deem it. For example, someone making a case for you in person often has an advantage over an email.
Overconfidence bias
This is exactly what it sounds like. You’re too confident in your own, or your team’s, expertise to give other ideas much consideration.
Feature positive effect
Focusing only on the positive benefits of an idea without weighing the negative impact it may have.
New ideas and different opinions can help you see that your team may be occasionally guilty of one of those 9—almost everyone is. However, once you realize this, you can be mindful of it in the future and take steps to prevent it.
